Cypher MAS

A hierarchical multi-agent system that detects hidden correlation patterns to collaboratively diagnose defects in industrial processes.

Python SPADE MongoDB Docker

About the project

Cypher MAS is a hierarchical multi-agent system designed to diagnose defects in industrial processes. Agents at different levels of the hierarchy collaborate to decipher hidden correlation patterns in process data, arriving at a collective diagnosis that would be difficult to achieve with a single model.

Stack used

  • SPADE as the multi-agent framework, providing the agent lifecycle and messaging infrastructure
  • MongoDB for storing process data, agent observations, and diagnostic results
  • Docker for containerising and orchestrating the agent system
  • Python for all agent logic, pattern analysis, and data processing